Bond yields rise, stocks don’t care #NewsMarket

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Good morning. Coke revenues fell a bit in the third quarter but, amazingly, they were still able to pass through double-digit price increases. Evidence that inflation is not truly dead — or that people just really like their sugar water? Email us: [email protected] and [email protected].

At some point, higher bond yields are supposed to matter for stock valuations

Those readers old enough to have been following markets in the bygone days of 2021 will remember that there is meant to be a strong, if not determinate, relationship between stock valuations and long-term bond yields. It makes sense, we were repeatedly told in those ancient times, that stocks valuations should be very high when Treasury yields are very low. Treasury yields are the discount rate used to derive stock prices from their expected future cash flows. This is especially true of fast-growing tech stocks, where a higher proportion of the expected cash flows are further in the future, which means the discount rate has more impact.

All of this makes perfect sense, so far as it goes. Discount rates do matter (though cash flows do too, and the two are related). In 2022, as the Federal Reserve tightened policy, the 10-year Treasury yield went from 1.5 per cent to 3.5 per cent and valuations fell as expected. The S&P 500 forward price/earnings ratio went from 23 to 16. Since then, however, the relationship has not been so well behaved. Treasury yields have traded sideways in a range between 3.5 per cent and 5 per cent. Valuations, meanwhile, have risen steadily; P/E’s are back in the mid-20s.

This is awkward for anyone who took the discount rate argument for high stock prices seriously back in the day. And it is getting more awkward still. Bond yields have risen from 3.6 per cent to 4.25 per cent in just five weeks. Shouldn’t stocks have the good manners to fall?

You can see how extreme the situation has become in this chart from UBS’s “Holt” research team, led by Michel Lerner. It shows the discount rate on US stocks (the rate that matches the market’s estimates for future free cash flows to their current price) against real US bond yields. The difference between the two (the red line) is the equity risk premium. The premium, at less than 2 per cent, hasn’t been lower since the dotcom madness 25 years ago. Their chart:

Holt metrics look forward, using market estimates. But you get a similar picture looking backward, using trailing average earnings. We can see this through Robert Shiller’s “excess cape yield”, which takes the cyclically adjusted earnings yield and subtracts the real Treasury yield. It is also less than 2 per cent and at the bottom of its historical range, though not quite as bad as 2001 and 1981. Here is Shiller’s chart, which also shows subsequent 10-year returns in excess of bonds (the green line):

At this point a certain subset of readers may be groaning that valuation levels have been a poor guide to subsequent returns for 15 years now. This is absolutely true. Since the great financial crisis, the only time high valuations have made themselves felt is 2022. But it can’t be that how much you pay for stocks will never again matter to the long-term returns they provide. That would be crazy.

One way to make the high price of equities less scary is to note that the composite profitability of the S&P 500 is extremely high, and suggest that this might be a more or less permanent condition because more of the index is composed of high-growth oligopolies: the Magnificent 7. Jason Trennert of Strategas describes this idea (about which he is a bit sceptical) as follows:

Profits margins for the S&P 500 are at all-time highs and the market’s return on equity is significantly higher than its 10-year average . . . One could be forgiven for thinking that we have reached some sort of new Olympian age in management skill . . . Of course, secularly low interest rates for 15 years can hide a multitude of sins . . . But for the time being the data would suggest that it is not at all surprising that the market is trading at historically rich valuations.

David Kostin’s strategy team at Goldman Sachs is not buying the Olympian Age thesis. They say the market’s great heights implies an expected real return of about one per cent a year over the next decade. Importantly, their model adjusts its projected returns upwards to reflect high current profitability. It estimates returns with a regression model that takes not just valuations, but also corporate profitability, recession probability, interest rates and, interestingly, stock market concentration. On this last point, Kostin argues:

Historical analyses show that it is extremely difficult for any firm to maintain high levels of sales growth and profit margins over sustained periods of time. The same issue plagues a highly concentrated index. Furthermore, the risk embedded in high concentration markets is not always reflected in valuation.

One could coherently argue this is “old economy thinking”: that Nvidia, Microsoft, Google and the rest have such good businesses that their growth and profitability will remain very high for decades on end. Certainly, these companies have defied expectations of normalising profits for many years already. They just don’t seem to regress to the mean.

All the same, it is important to remember what years of high returns make you forget: periods of very low returns, of the kind Kostin forecasts, are not all that unusual. Here is monthly data from Shiller showing 10 years of subsequent real S&P 500 returns for investing in any given month from 1870 to 2014:

Do you believe returns will never approach zero again, as they did in the 1910s, the 1930s, the 1970s, and the 2000’s?

One Good Read

Chutzpah.

FT Unhedged podcast

Can’t get enough of Unhedged? Listen to our new podcast, for a 15-minute dive into the latest markets news and financial headlines, twice a week. Catch up on past editions of the newsletter here.